The motivation of the “unified model” is to explain the main properties of the large zoo of active galactic nuclei with a single physical object. The discovery of broad permitted lines in the polarized spectrum of type 2 Seyfert galaxies in the mid 80’s led to the idea of an obscuring torus, whose orientation with respect to our line of sight was the reason of the different optical spectra. However, after many years of observations with different techniques, including IR and mm interferometry, the resulting properties of the observed dust structures differ from the torus model that would be needed to explain the type 1 vs type 2 dichotomy. Moreover, in the last years, multi-frequency monitoring of active galactic nuclei has shown an increasing number of transitions from one type to the other one, which cannot be explained in terms of the simple orientation of the dusty structure surrounding the active galactic nucleus (AGN). The interrelations between the AGN and the host galaxy, as also shown in the Magorrian relation, suggest that the evolution of the host galaxy may also have an important role in the observed manifestation of the nuclei. As an example, the observed delay between the maximum star formation activity and the onset of the AGN activity, and the higher occurrence of type 2 nuclei in star forming galaxies, have suggested the possible evolutionary path from, e.g., H II → AGN2 → AGN1. In the next years the models of unification need to also consider this observational framework and not only simple orientation effects.